Search

Artificial intelligence is providing new possibilities for analysis in the field of industrial radiography. As capabilities evolve, there is the need for knowledge concerning how to deploy these technologies in practice and benefit from the new automatically generated information. In this study, automatic defect recognition based on machine learning was deployed as an aid in industrial radiography of laser welds in an aerospace component, and utilized to produce statistics for improved quality control. A multi-model approach with an added weld segmentation step improved the inference speed and decreased false calls to improve field use. A user interface with visualization options was developed to display the evaluation results. A dataset of 451 radiographs was automatically analysed, yielding 10037 indications with size and location information, providing capability for statistical analysis beyond what is practical to carry out with manual annotation. The distribution of indications was modeled as a product of the probability of detection and an exponentially decreasing underlying flaw distribution, opening the possibility for model reliability assessment and predictive capabilities on weld defects. An analysis of the indications demonstrated the capability to automatically detect both large-scale trends and individual components and welds that were more at risk of failing the inspection. This serves as a step towards smarter utilization of non-destructive evaluation data in manufacturing. [ABSTRACT FROM AUTHOR]

Probability of detection (POD) as a metric for quantifying the capability of inspection procedures in nondestructive evaluation (NDE), has been applied and evolved in industries since 1970s. Progress had been noted when certain statistical functions were brought up to model POD behavior, including log-normal model (also referred as Probit model). This model had been concluded to be the best fit and therefore has been widely used in many studies, while the involved assumptions and conditions were not carefully addressed and explored. To make flexible POD datasets available for specific inspection procedures and reduce the number of expensive experiments needed, model-assisted POD (MAPOD) is an alternative. This paper addresses a pure simulation-based POD procedure of an inspection scenario involving phased array ultrasonic testing (PAUT) on lack-of-fusion defects in additive manufactured (AM) components. The mathematical simulations are performed by an ultrasonic testing (UT) simulation software, simSUNDT, developed at Chalmers University of Technology in Sweden. Resulted inspection datasets with the proposed data processing steps are evaluated in terms of the assumptions and conditions of log-normal POD model, with the purpose of discussing the POD model validity under different circumstances. Simulation-based POD curves are finally compared with several discrete POD values at some defect sizes, calculated through massive computations from physics-model based metamodel. Comparisons and observations confirm satisfactory application of log-normal POD model despite some violations in model hypotheses. [ABSTRACT FROM AUTHOR]

Eddy Current Testing (ECT) plays a key role in detecting cracks and defects in conductors. The present study examines for the first time how the subregion method as an effective mathematical and computational technique can be admixed with Finite Element Method (FEM) to study multiple defects parameters for ECT issues. Separating a defect region from the entire domain in any computational technique will save both time and storage space. Examples of different types of defects are presented in this article. A tangible result of processing time reduction by 90% has been achieved which has led us to consider the subregion FEM method as an effective method in solving different Nondestructive Evaluation (NDE) problems. An agreement between our results and others using classical FEM has been achieved which could lead to using this technique in online and field testing problems. The presented subregion FEM algorithm was verified experimentally with good agreement by testing Aluminum (T6061-T6) samples with defects. A Tunneling Magnetoresistive (TMR) sensor was used to measure the component of the magnetic field from normal to the sample top surface. A major component of minimizing processing time was achieved, which could lead to using this technique in online and field testing problems. [ABSTRACT FROM AUTHOR]

This paper presents an eddy current (EC) probe with orthogonal excitation coils that can change the exciting current direction electrically and tunnel magnetoresistive (TMR) sensor for inspection of stacked CFRP. As CFRP is electrically anisotropic, eddy current excited along the fiber direction is stronger than in other directions. A high sensitivity TMR sensor located at the center of the two orthogonal coils measures the z component of induced magnetic field (Bz) which constitutes the probe signal. The output of the sensor is 0 if the induced currents are symmetric about the sensor. Signals of defects in different orientations depend on the direction of the exciting current. The operation of the probe is validated using a finite element model. A prototype was built and experimental results from CFRP sample with defects in different orientations are presented and analyzed. [ABSTRACT FROM AUTHOR]

Eddy current methods are commonly used in the inspection of riveted multilayer structures. In a practical scenario, steel rivets may be magnetized to varying levels which can distort the magnetic field and pose a major challenge since the distortion can result in false positives and negatives. This paper presents an invariance transformation procedure that renders the signal insensitive to changes in magnetization while retaining the sensitivity to defect information. The approach is first implemented on simulated data, where the signals from magnetized steel rivets can be generated under controlled variation of relative permeability. The algorithm is then validated on experimental data acquired using steel rivets with different residual magnetization levels. [ABSTRACT FROM AUTHOR]

A computational technique is presented in this paper which improves finite element method (FEM) in solving inverse problems. Subregion method is used to select and isolate the area of interest where design parameters for defect shape are updated. An elastic mesh generator is developed in this paper to generate optimal meshes in the selected area to save connectivity matrix until having the most accurate design parameters. Using subregion FEM in solving inverse problems will help in minimizing processing time and memory usage in addition of reducing solution complexity. A 2-D eddy current testing (ECT) problem of detecting and characterizing the location and shape of a defect by separating the defect from entire domain is used to validate the presented subregion FEM algorithm. The elastic mesh generator is investigated to update the preselected design parameters of the defect in each iteration. This meshing technique adds the specialty of using subregion method in inverse problems where elements and nodes numberings are saved inside and outside the defect region. Both the genetic algorithm and simulated annealing optimization techniques are developed to get the accurate defect parameters. The presented subregion FEM results have been verified computationally using conventional FEM. Excellent results of signals agreement and processing time minimization with a factor of 90% with an accuracy of 98% have been achieved. In addition, the presented subregion FEM algorithm has been verified experimentally using aluminum (T6061-T6) and mild steel plate (0.15%–0.30% carbon and Fe) samples. This experiment is carried using an elongated excitation coil in a fixed position mounted on the top of the sample and tunneling magnetoresistive sensor to measure magnetic field. The measured magnetic fields were used as input to the inverse subregion FEM solver and the machined artificial defects were characterized with excellent accuracy. [ABSTRACT FROM AUTHOR]

The probability of detection (POD) of a well-controlled, automated eddy current (EC) procedure is evaluated in a numerical model and compared with experiments. The procedure is applied in laboratory environment with a single absolute probe which is positioned for raster scan over flat surfaces containing fatigue cracks. The variability of the signal amplitude, due to small fatigue cracks in the Titanium alloy Ti-6Al-4 V, is expected to mainly originate from crack characteristics and the index distance of the raster scan. The POD model is based on the signal versus crack size (versusa) result. The presented procedure provides a well-defined basis for a comparison between a simulated and an experimentally based POD assessment. Finite element analysis is used to model the EC method. A simplified fatigue crack model is first introduced and evaluated experimentally. Numerical computations are then used to build the corresponding model-based POD curve which shows good agreement with the experimental result. The model-based POD curve is generated both by means of a parametric and a nonparametric approach. Differences between model-based and experimental POD are discussed as well as the delta POD approach using transfer functions. [ABSTRACT FROM AUTHOR]